 |
|
|
|
 |
 |
|||
 |
|
|||
 |
|
September 2008 |
||
|
|
||||
| Clive Mott Director, Marketing and Strategy What trends are affecting the way users use your software? In general, the trend is away from a small number of functional experts, in fatigue analysis for example, to a much large number of engineers who undertake analysis across a range of technical areas, or designers doing “quick check” analysis. This demands that software comes preconfigured as an engineering application, not a tool kit. Appropriate settings and models should be preset up by the software house or an expert - as the staff running the applications do not always have a detailed background in all the technical aspects of the analysis. Test data files are getting larger –fast analysis is critical. Test data analysis will increasingly make use of parallelisation, or be done on central servers, and queued up. Analysis on the server also has the benefit of being able to store results from previous analysis, so the analysis is only run once, but used many times by different engineers. These trends point to ‘software as service’, although the volumes of test data, and commercial security implications seems to preclude it for most major players at the moment. The proliferation of engineers undertaking analysis increases the need to ensure corporate standards are maintained across geographically dispersed teams. One solution is to develop and publish ‘locked’ analytical processes which have been validated by the in-house expert. Again this tends to point towards networked based analytical solutions – for example, upload raw data; a standardised process takes place on the server, with the results published via web reports. There also is an increasing trend to ensure the product is optimised for the target market, rather than just sell product designed for one geography in another, with little or no engineering change. This leads to product better designed for the end use application (and lower warranty claims and associated costs). To do this effectively there is a need for better correlated test and operational data for particular user profiles – which will involve collecting more data from proving grounds, test fleets and real customer usage.
Software must increasingly demonstrate value, and be used further up the value chain in order to justify historic engineering software price points. The high end functionality of yesterday is often available in cheaper tools now, but the leading software developers are ensuring that software fits well into the product creation process, and the value is readily quantifiable. Point solution tools of the past are becoming fundamental parts the product creation process, and embedded in design processes – which all the communication and interfacing this demands. Often, users will have favoured analysis environments – e.g. Ansys, but not all the tools will be available from all suppliers. Increasingly specialist’s software companies (e.g. fatigue) will be embedded within these environments…. How is this affecting the way you will design the next generation of your software? nCode’s recently launched generation of software (ICE-flow 5) tries to address the trends noted above, and the future generations of software will add further support… The ICE-flow Automation software product from nCode analyses test and operational data on-line, and presents reports in a web environment. The full power of nCode’s analysis software (GlyphWorks) is incorporated into the server side of the software. Analysis processes can be set up to be run by anyone with web access, and reports are simply configured using a report builder tool. This product addresses locking down processes, engineering collaboration, and the collection and management of data from multiple test rigs and vehicle fleets. Real performance and test data (& derived information) is then ready to incorporate into the PLM process.
nCode are also looking towards constantly improving usability, but ensuring that the value of test and operational data is readily exposed and used across the design cycle. Basically this is about getting real usage and test data into the PLM process – something that exiting PLM systems are not designed to cope well with. (Large files sizes need specialist handling – something that nCode focuses on). In our CAE product range - we are working on ensuring that the underlying technology is open to every CAE environment – our own CAE fatigue product and interface is called DesignLife (and it is simple to use real test of operations data in the analysis), but we have built all our software in such a way so as to ensure that the underlying analysis engines can be used in other software environments. This allows users to use software environments they are familiar with, but still get access to the best in class technology. The general purpose FEA companies can offer specialist technology, but do not have to make the large investments needed to develop product. The general point is that software families increasingly have to be modular - allowing new functionality to be plugged in - new interfaces bolted on, and products can be rapidly interfaced. Also Enterprise level software (e.g. web based systems) should deliver valuable solutions with configuration - not development! Developing configurable software is more expensive than developing specific point solutions, but software companies will be able to address more applications and more quickly deliver value to the end users. Are there any simple ways you see for automotive engineers to improve the intelligence of their development processes? Many large OEMS often repeat tests and cae analysis. If the engineers could find relevant test and cae results rapidly, significant time and cost will be deleted from the product development cycle. Most CAE analysis systems are not designed to cope with large time series data sets, but this is now possible, giving a much better correlation between actual customer use and the analysis.
What’s holding your software’s ability back? Engineering software, by its nature, can be complicated. Increasing usage will continue to demand a focus on usability, and great interfaces, helping the walk engineers in a simple way through otherwise complex processes. There is always a need for more and more complete materials data, as well as fatigue models, especially for composites and exotic alloys etc. nCode will continue to make sure that new models and approaches are easily usable with the software framework we have developed. There is also a risk that the grant of patents for standard engineering approaches within software will hinder innovation – something that nCode will seek to avoid by aggressive innovation and moving the state of the art ever forward What are you doing to make the software more useful to engineers? Two main approaches: 1 – a continued emphasis on training - both via paid courses and free webinars and user groups. This helps to ensure that engineers fully appreciate the benefits of the analysis they are carrying out – and can see where and when to use the results. 2 – Building tools that are easily configured to the in-house process, rather that fixed approaches defined by the software house. The GlyphWorks and DesignLife products are ‘process centric’ tools – the end user designs the analytical process they need – and can then lock it and distribute the approach across the organisation. The ICE-flow Automation product deploys analytical processes can be deployed over the web, queries and reports can be defined centrally – and the software configured to meet a corporation’s work processes……. What is the next big step for your software? Web based collaboration will continue to be a focus – especially engineering analysis and large test and operational data sets. A common web-based system can expose different views on the same information – for use by different departments (and companies), helping consistency – but also cutting the cost of data handling. But crucially real test and engineering data can be easily incorporated into the design process. This is not just limited to fatigue and durability, (nCode’s analytical specialism) – but areas such as NVH, ride quality, and powertrain performance. How do you see the software industry evolving in the next five years? Software development cycles will not shorten dramatically – the entire development process of specification, developing test and documentation will still mean that most companies issue major release no more than once a year. In addition – many customers can not cope with more regular releases –as there is an overhead in deploying software across companies. However code quality should improve with the introduction of more automated testing techniques. This is often essential when seeking to stress test applications designed for large organizations and terabytes of data. Yes there will be more mergers – software companies will merge with software companies -looking to expand solution sets, also there will be more partnering – niche technology suppliers will work with the major players – and often all of them. Auto designers are likely to see fewer, but larger vendors , and specilist software comnpanies will often work through these larger companies. Software companies will also merge with non software companies – this will be more about vertical integration around particular engineering applications. These are often more challenging to make work, but done well can provide a powerful and focused solution set for key engineering sectors. |
nCode develops and supplies data analysis and data automation systems primarily focused on durability, fatigue and vehicle performance. The software is primarily used by OEMs and Tier Ones to analyse proving ground and test data, but also to carry out CAE fatigue analysis early in the design cycle – all aimed at helping vehicles and components meet durability targets, reduce weight and product costs. nCode ’s users are now adopting the concept of Product LifeCycle Performance – an approach that allows in - service operational performance data to be fed back into the design cycle. nCode develops web based data automation software that can acquire, manage, analyse and report on in - service data from operational, customer or test fleets. Europe is around half of nCodes user base - with around 1000 regular users of the software in Europe, mostly in the automotive sector, but also increasingly in the aerospace sector. LINKS Ansys: multi-physics analysis pays off. Read more... Integrated: Electromagnetic CAE tools that combine FEA and simulation. Read more... Lotus: vehicle dynamics made quick and easy. Read more...
|
|||
